Researchers use 'genomic pathway' to predict Parkinson's

Jun 17 2007

A new Mayo Clinic study provides strong evidence that the joint effects of common DNA variations in several genes that encode proteins within a well-defined biological pathway largely explain why some persons get Parkinson's disease while others don't, and even predict with great accuracy at what age people might develop their first symptoms.

The findings are published in the June 15 issue of PLoS Genetics.

"This represents a major paradigm shift from single gene studies to genomic pathway studies of complex diseases," says Demetrius Maraganore, M.D., the Mayo Clinic neurologist and Parkinson's disease specialist who led the study.

The authors say traditional genetic studies have either discovered rare single gene mutations that cause Parkinson's disease only in isolated families, or have identified common single gene variants that are only weakly associated with the disease and inconsistently across populations.

"By examining a large cluster of related genes, we found patterns that make people up to 90 times more likely to develop Parkinson's than the average person," says study co-author Timothy Lesnick, a Mayo Clinic biostatistician. "The size of the effects that we observed for genes within a pathway and the statistical significance of the predictive models were unprecedented."

The models were highly effective in predicting age of onset of the disease: by age 60, 91 percent of patients in the highest-risk group already had Parkinson's, while only 11 percent of patients in the lowest-risk group did. By age 70, every member of the highest-risk group had the disease, whereas two-thirds of patients in the lowest-risk group still were disease-free. Members of the highest-risk group typically developed Parkinson's more than 20 years earlier than the lowest-risk group.

The researchers speculate that common genetic variations within the same biological pathway might also contribute to a person's risk of developing other brain diseases; disorders such as Alzheimer's disease, Tourette's syndrome, dyslexia, epilepsy and schizophrenia, need to be studied.

The study is significant because:

1. The genetic strategy was novel, investigating not just one, but many genes that were predicted to interact in a specific biologic pathway. This study provides a blueprint for the study of other complex diseases, which often have only modest associations with certain single genes. The summation of small effects from many genes in the same biologic pathway may be key to understanding many human diseases including Alzheimer's, cardiovascular disease, diabetes, nicotine and alcohol dependence, and many cancers, in addition to Parkinson's disease.
2. It provides intriguing new insights into the symptoms of Parkinson's disease, and may lead to tests to identify persons at high risk and to new treatments to prevent the disease or halt its progression.
3. These high-risk Parkinson's disease genes were contained in the biologic pathway for the development of the human brain, as well as repair and remodeling of brain circuits (so-called axon guidance pathway). This raises speculation for another environmental influence that heretofore has been ignored: pre-birth events in the mother's womb.

The researchers studied the axon guidance pathway, which includes at least 128 genes that encode proteins that play a critical role in guiding the axons (or wiring) of the brain during fetal development. The same proteins also repair the wiring and determine the fate of damaged brain cells later in life.

The Mayo researchers analyzed a dataset that included information for hundreds of thousands of common DNA variants (single nucleotide polymorphisms or 'sNPs) in 443 Parkinson's disease patients and in 443 control subjects; the controls were unaffected siblings. Researchers identified SNPs within axon guidance pathway genes and used statistical methods to identify combinations (models) of SNPs that were highly predictive of susceptibility to Parkinson's disease, survival free of Parkinson's disease, and age at onset of Parkinson's disease. They then validated their findings using data from additional whole genome DNA and RNA variation datasets for Parkinson's disease.

Dr. Maraganore and Mayo colleagues published the first whole-genome association study of Parkinson's disease in 2005, studying hundreds of thousands of common DNA variants in the human genome, but as single and unrelated risk factors. They identified a dozen DNA variants that were weakly associated with Parkinson's disease, but those findings have since been difficult to replicate. Their most significant finding was for a DNA variant within the axon guidance pathway. This was the key clue in that first study, leading them to predict that this axon guidance pathway needed to be investigated. Thus, they focused on a few thousand variants within genes that encode that pathway, and analyzed their joint effects.

Dr. Maraganore compares the change in method to the difference between gazing at faint stars with a telescope versus looking at the Milky Way with the naked eye. "Instead of straining to observe the effects of single DNA variants scattered randomly across the universe of the human genome, we observed the effects of a constellation of variants within a well-defined genomic pathway," he says. "When we looked at the axon guidance pathway in Parkinson's disease in this way, it lit up."

"The effect of any single gene variant was small and of questionable significance, but the joint effects of multiple gene variants within the pathway were profound," Lesnick explains. "Not only did we identify persons with as much as a 90 times greater risk for Parkinson's disease, but we determined that the odds of our findings occurring by chance were less that one in a trillion trillions." The findings also were validated in two independent datasets.

Still, the researchers caution that the predictive model findings need to be tested across global populations. They add that the axon guidance pathway should be studied in other diseases, and that other pathways should be studied in Parkinson's disease, to add context to their findings. They also say mechanistic studies in experimental models will be required to understand precisely how variation in the axon guidance pathway leads to Parkinson's disease.

The researchers say this is among the first times that information from a dataset of DNA variations across the whole genome has been employed to study the joint effects of a subset of DNA variations within a defined biological pathway (genomic pathways" approach).

"This study serves as a blueprint for how we can study all complex disorders. By studying genomic pathways, we might be able to determine what causes people to be obese or to have hypertension, or to develop diabetes or heart disease," Dr. Maraganore says.

Implications for Patients

Parkinson's disease is an aging-related disorder that affects nearly a million Americans. The cardinal signs include tremors, slowness of movements, and rigid body and limbs. The symptoms predominantly arise from progressive degeneration of brain cells that produce the chemical dopamine. "It is intriguing to consider that patients with Parkinson's disease may be wired differently from birth," adds co-author Eric Ahlskog, M.D., Ph.D., the head of Mayo Clinic's Movement Disorders Section. "This might explain lifelong differences in personality, including avoidance of addictive substances and increased anxiety and depression. It also might explain asymmetries in the movement disorder that occurs in Parkinson's disease; typically, one side of the body is affected predominantly."

The findings could quickly lead to genetic tests to identify persons with a high probability of developing Parkinson's disease during their lifetime. For example, a person with high predicted probability could have a risk of Parkinson's disease that is 90 times greater than the average person; members of intermediate-risk groups would be four to 25 times more likely to develop Parkinson's. The findings may also lead to development of new treatments that would promote the repair of damaged axons in the brain and spare nerve cells from early death, which could prevent or slow the progression of the disease.

Researchers say that just as subtle genetic variations within the axon guidance pathway might alter brain wiring during fetal development and predispose to Parkinson's disease decades later in life, subtle maternal health factors also could alter brain wiring and predispose to the disease.

The researchers say effects of maternal health on the risk of Parkinson's disease should be explored in future studies. Until now, environmental studies of Parkinson's disease have largely focused on adulthood exposures, such as to pesticides, smoking and coffee. These same exposures might contribute to the cause of Parkinson's disease during gestation, they suggest.

http://www.mayoclinic.org/